(T-041) Bimekizumab population pharmacokinetic and exposure-response analyses in adults with moderate to severe hidradenitis suppurativa

Pinky Dua, PhD – Global Head of Quantitative Clinical Pharmacology and Pharmacometrics, UCB Pharma; Siv Jönsson, PhD – Senior MIDD consultant, Pharmetheus AB; Julie Janssen, PhD – MIDD consultant, Pharmetheus AB; Akash Khandelwal, PhD – Global head of Pharmacometrics, UCB Pharma; Bob Rolleri, PhD – Executive Director, Clinical Development, UCB Pharma; Jagdev Sidhu, PhD – Head of Ethnopharmacology, UCB Pharma

Objectives: Population PK and ER analyses to support Bimekizumab (BKZ) clinical dosing in adults with moderate to severe hidradenitis suppurativa (HS).

Methods: Longitudinal PK (1035 patients, 8891 concentrations, average 8.6 per patient) and HS clinical response (HiSCR [observed case], 997 patients, 11942 observations, average 12 per patient) data from 3 HS studies (BE HEARD I/II, HS0001) were used to develop the models in NONMEM. Treatments in HS0001: 640 mg BKZ loading dose followed by 320 mg BKZ Q2W over 12 weeks; and in BE HEARD I/II: placebo, 320 mg BKZ Q2W or Q4W to Week 16, followed by 320 mg BKZ Q2W or Q4W maintenance to week 48 (denoted placebo/Q2W, Q2W/Q2W, Q2W/Q4W, Q4W/Q4W). Patients were 57% females, 59%/41% Hurley stage II/III with median (range) age and weight (WT) of 35 (18–78) years and 95 (48–196) kg.

Results: BKZ PK was described as a 1-compartment model with 1st-order absorption and elimination with baseline WT, prior use of biologics and disease severity as covariates. For a patient with Hurley stage II, weighing 95 kg and no prior biologics use, CL/F was 0.512 L/day, V/F 14.5 L, ka 0.499 day-1 and half-life 19.7 days.

A proportional odds model described the HiSCR data and the probability of being a non-responder and HiSCR50/75/90/100 responder (50/75/90/100% reduction in the total abscess and inflammatory nodule (AN) count with no increase from baseline in abscess or draining tunnel count) was a function of baseline probability, treatment, previous HiSCR observation and inter-individual variability. The probability of response increased over time and with increasing BKZ concentration (Emax function). The EC50 was 0.956 µg/mL, and the derived EC90 (8.60 µg/mL) is lower than observed pre-dose concentrations at Week 16 in 80% of the patients, illustrating that with the studied treatments most patients were close to Emax. The HiSCR response probability (all thresholds) was lower in smokers vs. non-smokers and decreased with increasing baseline AN count.
The median predicted (observed) proportion of HiSCR50 responders at Week 16 were 62% (61%), 59% (59%) and 62% (59%) following 320 mg BKZ Q2W/Q2W, Q4W/Q4W and Q2W/Q4W, respectively. At Week 48 the responses were 82% (77%) (Q2W/Q2W), 81% (79%) (Q4W/Q4W) and 80% (81%) (Q2W/Q4W). Simulations predicted non-smokers to have ≈10% higher response rates compared with smokers, and patients with high baseline AN count (≥20) to have ≈5–10% lower response rates compared with those with low ( < 20) baseline AN count, findings independent of dosing regimen.

Conclusions: In this work, representing the first reported pharmacometrics analysis of HS efficacy data, similar HiSCR outcomes were predicted during initial and maintenance treatment periods for BKZ 320 mg Q2W/Q2W, Q2W/Q4W and Q4W/Q4W dosing, and indicated that no specific sub-group (smoker, high baseline AN count, high WT, Hurley stage III, prior biologics use) would benefit from a different dosing regimen than the rest of the HS population.

Citations: Not applicable